Re: [PATCH v2 3/3] fetch: reduce duplicate in ref update status lines

Nguyễn Thái Ngọc Duy  [off-list ref] writes:

+static int common_suffix_length(const char *a, const char *b)
+{
+	const char *pa = a + strlen(a);
+	const char *pb = b + strlen(b);
+	int count = 0;
+
+	while (pa > a && pb > b && pa[-1] == pb[-1]) {
+		pa--;
+		pb--;
+		count++;
+	}
+
+	/* stick to '/' boundary, do not break in the middle of a word */
+	while (count) {
+		if (*pa == '/' ||
+		    (pa == a && pb > b && pb[-1] == '/') ||
+		    (pb == b && pa > a && pa[-1] == '/'))
+			break;
+		pa++;
+		pb++;
+		count--;
+	}
+
+	return count;
+}
+

Why do you need two loops, one going backward from the tail and then
going forward toward '/'?  Wouldn't it be sufficient to keep track
of the last slash you saw in a while scanning backwards?  I.e
something along the lines of:

	tail_a = a + strlen(a);
	for (pa = tail_a, pb = b + strlen(b), slash_in_a = NULL;
             a < pa && b < pb && pa[-1] == pb[-1];
	     pa--, pb--) {
		if (*pa == '/')
			slash_in_a = pa;
	}
	count = a + strlen(a) - slash_in_a;
		
perhaps?

These are somewhat cryptic ;-)

Other than that, the patch looks OK.

Thanks.